A metric is a measured property of a system.
Examples of a metric include:
the response time of an application
the utilization of a CPU
the amount of time spent in garbage collection
the cost of a cloud service
Metrics are used to both specify the optimization goal and constraints (e.g. minimize the heap size while keeping response time < 1000 and error rate <= 10% of a baseline value), and to assess the behavior of the system with respect to each specific configuration applied.
A metric is described by the following properties:
a name that uniquely identifies the metric
a description that clarifies the semantics of the metric
a unit that defines the unit of measurement used by the metric
The construct to be used to define a metric is described on the Metric template page.
Metrics are displayed in the Akamas UI when drilling down to each system component.
and are represented in metric charts for each specific optimization study
Please notice that in order for a metric to be displayed in the Akamas UI, it has to be collected from a Telemetry Provider by means of a specific Telemetry Instance defined for each specific target system.
